AI Slop by the Numbers: What's Actually Happening to Your Feeds (2026)

"AI slop" went from niche complaint to dictionary-adjacent phrase in about two years. But beyond the vibes — the uncanny thumbnails, the narrated "fact" channels, the synthetic cooking videos — what does the data actually say about how much AI-generated content is in your feeds, and how people feel about it?

Short answer: it's a lot, people have noticed, and sentiment has fallen off a cliff. Here are the numbers, with sources.

How much AI slop is actually in your feed?

The most striking figure comes from a late-2025 study of YouTube's recommendation system: roughly 21% of recommendations served to new users were AI-generated slop, with about 33% more classified as low-effort "brainrot" content (financialcontent.com). For a brand-new account with no watch history, that means more than half of what the algorithm offers up may be either synthetic or junk.

Users' own reports line up with that. In EMARKETER's survey research, 56% of people say they see AI slop on social media often or very often, and 83% see it at least sometimes (emarketer.com). This isn't a fringe experience anymore — it's the default condition of scrolling.

The sentiment collapse: 60% → 26% in two years

If you only remember one statistic from this article, make it this one.

In 2023, 60% of consumers said they were enthusiastic about AI-generated content from creators. By 2025, that number had fallen to 26% (emarketer.com).

That's not a dip; that's a collapse. The novelty period — when an AI-generated image or voiceover was a curiosity — is over. Audiences have seen enough of the median output to form an opinion, and the opinion is largely negative.

Trust is eroding alongside enthusiasm

Two more data points sketch the trust picture:

• About 62% of consumers say they are less likely to engage with or trust content they know was AI-generated (sproutsocial.com).
• Gartner found 53% of consumers distrust AI-powered search results (gartner.com).

Put together: a majority of people see slop regularly, a majority discount content once they know it's synthetic, and a majority don't trust AI-mediated search either. The default posture toward AI content has shifted from curiosity to skepticism.

The platforms know — and they're saying so out loud

This isn't just users grumbling. YouTube CEO Neal Mohan publicly named managing AI slop and detecting deepfakes as a priority for 2026 (cnbc.com).

That admission matters for two reasons. First, it confirms the scale of the problem from the inside. Second, it signals that platform-level fixes — disclosure labels, demonetization of mass-produced content, deepfake detection — are coming, but on the platform's timeline, optimized for the platform's incentives. If you want a cleaner feed today, the platforms aren't going to hand you one.

The infrastructure behind the flood

To understand why the volume of AI content keeps climbing, look at what's being built to produce it.

The International Energy Agency projects that global data centre electricity consumption will roughly double from about 485 TWh in 2025 to about 950 TWh by 2030 — approaching 3% of all global electricity — with AI as the most significant driver. Demand from AI-optimised data centres is projected to more than quadruple over that period (iea.org; datacenterdynamics.com; nature.com). The IEA also expects about half of that demand growth to be met by renewables, with natural gas expanding by roughly 175 TWh (iea.org).

The takeaway for your feed: generation capacity is scaling up massively, and generating content is the cheapest it has ever been. The flood is not a temporary glitch. The economics guarantee more synthetic content next year than this year.

An honest note before anyone overreads this: blocking AI content from your feed does not reduce AI energy consumption or emissions. The content is already generated by the time it reaches your screen; filtering it changes what you see, not what gets produced. If a tool implies that hiding slop is climate action, be skeptical. The energy numbers here are context for why the volume keeps growing — the real stakes for an individual are attention and trust, not carbon.

What you can actually control: your own feed

You can't stop the flood at the source. You can decide how much of it reaches you. Realistic options in 2026:

1. Train the algorithm manually. Use "Not interested" and "Don't recommend channel" on YouTube, hide and snooze on Facebook. It works, slowly and leakily — new slop channels appear faster than you can dismiss old ones.

2. Use a slop-filtering browser extension. A small ecosystem has emerged here, and you have genuinely free options: Clarity, ByeAI, and AI Slop Blocker are free, YouTube-only filters. AI Content Shield covers multiple platforms on a freemium model (around $59/year for the paid tier).

3. Unslop (our tool, so apply appropriate salt) sits in between: it filters YouTube — home, search, sidebar, and Shorts shelves — plus the Facebook main feed. It works by scanning visible text, hashtags, your own custom keywords, and the platform's "Altered or synthetic content" disclosure label, with whole-word matching so "AI" doesn't accidentally nuke videos about Dubai, email, or rain (there's a 16-test suite specifically for that). It runs 100% locally — no account, no server, no telemetry, just the storage permission — and the core version is free with 20 custom keywords. Pro is a one-time $5 unlock (pay-what-you-want from $3), not a subscription, verified offline.

Worth stating plainly what text-based filters — Unslop included — can't do: they don't analyze pixels or audio, so undisclosed AI content with clean titles can slip through, and none of them detect AI voices acoustically. They catch the labeled, the hashtagged, and the keyword-obvious slice of slop — but not all of it.

The bottom line

• 21% of YouTube recommendations to new users are AI slop; ~33% more are brainrot (source)
• 83% of people see slop at least sometimes; 56% see it often (source)
• Enthusiasm for AI creator content: 60% → 26% in two years (source)
• 62% trust AI-generated content less (source); 53% distrust AI search (source)
• The infrastructure to generate more is doubling (source) — and platforms admit the problem (source)

The flood isn't slowing down. The question is no longer whether AI slop reaches your feed — it's whether you decide what stays in it. Whether you do that with the platform's built-in tools, a free extension, or something like Unslop, the data says it's worth doing deliberately.